- Email: [email protected]
Diazepam alters brain-stimulation reward thresholds in seizure-prone sites
Behavio,tral Brahl Research, 46 (1991) 167-173 9 1991 Elsevier Science Publishers B.V. All rights reserved. 0166-4328/91/S03.50
167
BBR01242
Diazepam alters brain-stimulation reward thresholds in seizure-prone sites T. Harris and C. Bielajew School of Psychology, University of Ottawa, Ottawa, Ontario (Canada)
(Received 8 January 1991) (Revised version received 8 July 1991) (Accepted 21 September 1991) Key words: Benzodiazepine; Diazepam; Brain-stimulation reward; Self-stimulation; Lateral hypothalamus; Ventral tegmental area; Lateral preoptic area; Seizure
Studies of the effect of diazepam and related compounds on the rewarding properties of brain stimulation as measured by response rates have not yielded clear results, with self-stimulation performance reported to be potentiated, diminished, or unchanged folloMng drug administration. In this study, the effect of two doses ofdiazepam (2.5 and 5.0 mg/kg) and its vehicle on self-stimulation thresholds was examined in eight rats with electrode placements scattered along a 4 mm length of the medial forebrain bundle. Stimulation of the lateral preoptie area and the anterior and mid-lateral hypothalamus produced overt seizures. Rate-period curves were generated for a wide range of currents and the resulting period-current trade-off functions were compared across doses. In seizure-prone sites upward shifts in period threshold were observed after 2.5 mg/kg ofdiazepam with little additional increases incurred by the 5.0 mg/kg dose. The majority ofnon-seizure sites showed no effects ofdiazepam upon period threshold. The results suggest that diazepam alters brain-stimulation reward thresholds by suppressing competing seizure activity.
INTRODUCTION The benzodiazepines constitute a class o f m i n o r tranquilizers that have anti-convulsive, sedative, and antianxiety properties 9. Reports o f the effect o f these c o m p o u n d s on brain-stimulation reward in the lateral hypothalamus ( L H ) have not been consistent II. Some studies have found a facilitation o f self-stimulation response rates at low doses (0.5-10 mg/kg) of benzo~ diazepines ~'4'15 while others have found either no effect or a decrease in rates over the same range o f doses 3"4. Likewise with higher doses (10-120 mg/kg), both facilitated and depressed response profiles have been reported despite observed low dose elevations in rates in the same animals ~s. These equivocal results may be attributable to an unpredictable interaction between the sedative actions o f benzodiazepines at high doses and the effects o f brain stimulation, by either promoting rates through suppression o f aversion in mixed placements or reducing rates due to sedation in pure-reward sites.
Correspondence: C. Bielajew, School of Psychology, University of Ottawa, Ottawa, Ontario, Canada, KIN 6N5.
The influence o f benzodiazepines on extrahypothalamic placements has not been extensively studied. Caudarella et al. 4 found that low doses o f diazepam facilitated L H response rates while dorsal hippocampal rates were attenuated. In the prefrontal cortex, response rates were unaffected if d i a z e p a m was administered to animals following acquisition o f self-stimulation behaviour at. A major d r a w b a c k o f these studies is the pervasive use o f response rate as an index o f reward; the assumption that drug-induced changes in rate indicate a shift in reward is unfounded. T h e performance enhancing or debilitating effects o f some drugs may mask the 'true' reward effects 5,6,1~ T h e threshold method for assessing changes in self-stimulation controls and minimizes the contribution o f performance factors. Typically, current is held constant while the frequency o f pulses, or its reciprocal, period, is systematically varied to generate what has been termed a 'reward summation' function 5, that is, a r a t e - f r e q u e n c y or r a t e - p e r i o d curve. Threshold is calculated by interpolating the frequency or period corresponding to a criterion response rate. Because absolute response rates are ignored, the problems ofinterprctation due to non-motivational fac-
168 tors are eliminated and the reliability of the data strengthened. Very few studies have evaluated threshold changes in self-stimulation following benzodiazepine administration. In the prefrontal cortex, current thresholds are unaltered by diazepam challenge2% while in the LH, the frequency thresholds appear to decrease in subjects with pure reward placements; this finding is less consistent in subjects with mixed reward-aversion sitesL To our knowledge, increased current or frequency thresholds have not been observed, suggesting that the benzodiazepines, when effective, amplify the rewarding properties of electrical stimulation. This notion is strengthened by the finding that administration ofpicrotoxin, a benzodiazepine antagonist, results in increased frequency thresholds, implying a facilitation of reward iT. Finally, diazepam has been reliably shown to produce a significant conditioned place preference ~4"22, consistent with the notion that diazepam has rewarding effects. However, the question still remains as to why the benzodiazepines are not effective in altering self-stimulation thresholds at all sites, and in all subjects at the same site. One suggestion is that diazepam's rewardenhancing effects are reduced in placements with aversive properties ~'2. If so, this phenomenon seems to be restricted to the hypothalamic region as reward thresholds in sites not usually associated with aversion (e.g. prefrontal cortex) are basically unaffected by diazepam 2~. A second possibility, one that emerged from our pilot data, is that the anti-convulsive action of benzodiazepines suppresses the development or appearance of motor seizures that inhibit the behavioral expression ofthe rewarding effects of the stimulation; thus, in those animals not experiencing seizures, self-stimulation thresholds would be unaffected by the anti-convulsant property of the drugs. This interpretation is difficult to assess in earlier studies because seizure activity was not reported. Thus, the present study was initiated to determine if threshold changes were correlated with the presence of seizures. Electrodes were implanted along a wide posteroanterior axis of the medial forebrain bundle from the ventral tegmental area (VTA) to the lateral preoptic area (LPO) to include sites where the probability of observing stimulation-induced seizures was high.
MATERIALS AND METHODS
Subjects attd surgery Eightmale Long-Evans rats (Charles River Laboratories) were individually housed in plastic cages under
a 12 I1 light/dark cycle with ad libitum access to food and water. At the time of surgery, the individual weights ranged from 300 to 440 g. The rats were anesthetized with sodium pentobarbital (65 mg/kg i.p.) and xylazine (5 mg/kg i.m.). A subcutaneous injection of atropine sulfate was administered to reduce respiratory distress. Standard stereotaxic procedures were used to implant the electrodes in one of three sites using a fiat-skull orientation ~6. The three VTA subjects had coordinates ranging from 4.3 to 4.8 mm posterior to bregma, 0.9 to 1.2 mm lateral to the mid-sagittal suture, and 8.4 to 8.8 mm below the skull surface reading at bregma; the four LH subjects had coordinates ranging from 2.0 to 3.0 mm posterior, 1.6 to 1.7 mm lateral, and 8.2 to 8.4 mm ventral; the one LPO subject had coordinates of 0.6 mm posterior, 1.3 mm lateral, and 8.2 mm ventral. The electrodes were fashioned from stainless-steel wire, 250 pm in diameter, and insulated with Epoxylite to the flat tips. The indifferent electrode consisted of a 2-56 stainless-steel machine screw with a stainless-steel flexible wire soldered to it; the wire was wrapped around the four stainless-steel skull screws. The entire assembly was anchored to the skull screws by dental cement.
Apparatus The testing chamber was a 28 cm long x 38 cm wide • 44 cm tall wooden box with a Plexiglas front. A rodent lever protruded from a side wall and was positioned 3 em above the floor. Stimulation was delivered by a constant-current amplifier ~3 and an integrated circuit pulse generator, built in-house. The current was continuously monitored on an oscilloscope by reading the voltage drop across a 1 kf~ precision resistor in series with the rat. Each lever depression delivered a 500 ms train of rectangular monophasic cathodal pulses of 100#s duration. The pulse period and current were varied as described below.
Drugs Diazepam (Hoffman-LaRoche) was obtained in injection ampules at a concentration of 5 mg/ml. The vehicle solution was prepared from 414 mg propylene glycol and 80 mg ethanol per ml of saline, having a pH of 7.3, and was filtered before each injection.
Screen#lg and stabilization Animals were allowed to recover from surgery for at least 3 days before screening for self-stimulation began. Standard operant shaping procedures were used to train the rats to bar-press for electrical stimulation.
169 Once reliable performance was observed, the onset of each 60 s trial was signalled by 5 stimulation primes with current and period identical to the 60 s trial stimulation; the inter-prime interval was 1.0 s. The current and period were adjusted to produce rates of responding greater than 50 per minute. During stability sessions, an ascending series of periods was introduced with the initial period sufficient to produce high rates of responding and each successive period a 0.10 common log unit step above the previous one. The series was continued until the rate of responding in a 60 s trial was 5 bar presses or fewer. This procedure was applied to 6 or 7 currents which were spaced roughly equally from 80 ILA to 1000/~A. Rate was plotted against period for each current to determine the threshold, which was defined as that period corresponding to 25 responses and was interpolated from the rate-period functions shown in Fig. 1. From this family of curves, a period-current trade-off function was generated. Two measures of stability were used to determine when a subject was ready to begin drug tests. Within a session, a current, usually 200 ltA, and its associated period threshold was designated as the baseline measure. If the baseline threshold measured at the beginning and at the end of the session differed by no more than 0.10 common log units, the data from that session were retained; if greater, the session was discarded. The second stability measure was a between-
foe 9r
I00 8C ~
.~
160
250
5OO \ LA I000
7C
session criterion. The standard error of the mean associated with the average period threshold across sessions could not exceed 10~ of the mean value. Replications of the full period-current trade-off function were continued until the majority of the average period thresholds met this criterion; typically this required from 4-6 sessions, with a single period-current trade-off function generated per session. A session lasted approximately 2 h. Once the between-session stability was established, drug testing began.
Behavioural testhlg Six of the eight subjects were tested with two doses of diazepam, 2.5 mg/kg and 5.0mg/kg, while the remaining two subjects (URF62 and F472) were tested with only the 2.5 mg/kg dose. All subjects received vehicle injections. During test sessions, the incidence of stimulation-induced seizures or signs indicating the development of seizures was recorded. Drug and vehicle days were interdigitated so that drug tests were conducted no less than 48 h apart. On those days, the two doses ofdiazepam were assigned randomly. Within each session, the currents were also tested on a random schedule to yield a full period-current trade-off curve. The between-session criterion mentioned above was imposed during this phase of the study; thus, 4-6 replications of the entire period trade-off function were collected under each drug condition. The subjects were injected intraperitoneally approximately 10 min before testing began. This delay allowed the overt signs of the sedation induced by diazepam to subside. The duration of the test sessions was roughly one hour, during which time, stable baseline measures were usually observed.
Histology Subjects were perfused transcardially with a solution of 0.9~o saline followed by 10~ formalin. The brains were removed from the skull and stored in 10~o formalin for at least 48 h before sectioning and staining with Cresyl violet. The Paxinos and Watson atlas ~6was used to verify the location of the electrode tips.
i
\ 5O
... 2C
RESULTS 4.0
""
~3
~ao
I~e
zs.~
3s.
63.1
I00.0
PERIOD THRESHOLD (msec)
Fig. I. An example of period threshold determination from r a t e - period functions collected at different currents. A constant criterion of 25 responses/min was used to evaluate period threshold as indicated by the horizontal dashed lines. The corresponding thresholds which are indicated by the downward arrows were used to derive the period-current trade-off functions presented in Fig. 3.
Histology Tracings of the atlas plates ~6 corresponding to the sections containing the electrode tips are shown in Fig. 2. Subject TH22's electrode was located within the limits of the LPO, slightly ventral and medial to the substantia innominata. Of the four LH electrodes, three of the tips were found at different depths in the same
170
TH22
F472 oURF62 mTH21 ATH20
THIS
THI9
THII
Fig. 2. Tracings from atlas plates (Paxinos and Watson, 1986) that best correspond to the sections containing the electrode tips. The anteroposterior level is shown at the bottom of each tracing. The alphanumeric listing on the right of each section identifies the subject. Electrode placements ranged from the posterior LPO to the posterior VTA.
anteroposterior level. One tip (TH20) was positioned between the dorsomedial hypothalamic nucleus and the perifornical area, just dorsal to the level of the fornix. Another (TH21) had a similar placement but was marginally more dorsal and lateral, just below the subincertal nucleus. The last tip located at this level (URF62) was in the zona incerta just lateral to the mammillothalamic tract. Subject F472's electrode was situated far more anterior to the other LH placements with the tip in the most lateral part ofthe LH just dorsal to the magnocellular preoptic area. The three VTA electrodes were found at different anteroposterior levels. The most anterior (TH18) was at the dorsal border of the VTA, slightly lateral to the fasciculus retroflexus. The second (TH 19) was just ventral and medial to the medial lemniscus, and the most posterior VTA placement, (TH 11), was positioned at the ventral border of the VTA encroaching on the medial lemniscus and paranigral nucleus. Dntg Tests
The effect of diazepam upon brain stimulation reward is shown in Fig. 3.
The subjects are presented according to their anteroposterior electrode placement, from left to right beginning at the top, so that the data from the subject with the most posterior electrode placement (TH 11) appears in the top left box and the data from the subject with the most anterior electrode placement (TH22) appears in the bottom right box. The individual curves represent the average period thresholds plotted against current (period-current trade-off function) that were obtained under each drug condition with vehicle tests distinguished by filled circles and solid lines, 2.5 mg/kg tests distinguished by filled circles and dashed lines, and 5.0 mg/kg tests distinguished by open circles and solid lines. Seizures which ranged from head and mouth movements with wet dog shakes to fully expressed motor seizures with rearing and falling were observed repeatedly during vehicle testing in TH22, F472, and TH20. Administration of diazepam eliminated or reduced the severity of the seizures. The remaining five animals showed no overt seizure signs. Stimulation of the five seizure-negative sites resulted in little or no effect of diazepam upon self-stimulation thresholds at either the 2.5 mg/kg or 5.0 mg/kg dose. However, the three seizure-positive placements showed a striking increase in period thresholds, ranging from 0.1 to 0.4 common log units, following diazepam challenge. In these three subjects, threshold shifts were observed with the 2.5 mg/kg dose; the higher dose yielded little or no additional increase in period thresholds. The total charge was calculated for each of the individual period-current trade-off functions using the equation Q = INd
where Q is the charge in ILC, I is the current in/tA, N is the number of pulses in the stimulation train, and d is the pulse duration in seconds 8. The charge value across currents should remain constant if there is an equal reciprocity between period and current; if so, the slope of the period-current trade-off function that relates these two variables, when plotted using logarithmic (base 10) scales, should be 1. Although there is a wide range of slopes across subjects in this study, the slopes of the trade-off functions within subjects are similar, implying that equivalent changes are occurring at each current across drug tests in individual subjects. The use of charge simplifies the data analysis by combining the information conveyed in the trade-off curves into one number which represents the entire curve. Therefore, for every one of the 4-6 replications of each
171 158.5 1259
HII
I000
f
79.4
THI9 -5.3
THIS
URF62 -31
TH20 -5.1
F472
TH22 -0.6
6~1.1 50.1 398 31.6 25.1 20.0 "~
158 126 I00
7.9 6l]
::I:: 4o ~ 3.2 1~5.9 RO0.O
79.4
TH21 -3.1
;e
-I.8
.~./.
seizures
63.1
seizures
seizures s"
I~.
39.8
_ 4 . "~
p.-
31.6 25.i
S/
20.0 158
se~
12.6 I00
T.9 6.3 5.0 4.0 3240
~;
l ,GO ' '~0 ' z~O ' 4;0
~iO ' '0;0
9 6~ ',oo " ,6o =~o " ,Go" 6~o ,ooo
63
CURRENT
,oo ' ,~o " =~o
,Go
6~o " ,o'oo
9 s'3 ' ,6o"i~o
=5o',bo'e~o'lo'oo
(p-A)
Fig. 3. The average period-current trade-off functions determined for all drug conditions with period threshold plotted against current. Each line is based on 4-6 replications. In the upper left corner of each box is the subject's identification number and below that, the anteroposterior coordinate for that subject's electrode placement. Filled circles joined by a solid line represent the vehicle condition; filled circles joined by a dashed line represent the 2.5 mg/kg diazepam condition; open circles joined by a solid line represent the 5.0 mg/kg diazepam condition.
trade-off function shown in Fig. 3, Q was calculated for every data point, that is, each period threshold value that contributed to the curve. The charge values were
~7
summed across the currents.to yield a total charge for each replication which was then averaged to yield one representative value for that test condition (Fig. 4).
~,b,c
a-O vs 2.Srng/kg p
46 btc
.7-
.Z.J.
J. _z. dr..
E 3 t.t.
0 2.5 5.0
THII
2~ 5,0
THI9
0 2.5 S.O
0 2-5
0 2.5 SD
THI8
URF62
TH21
Z.5 5.0
TH20
0 25
F472
1
0 2~5S0
TH22
DOSE OF DIAZEPAM (mg/kg)
Fig. 4. The graph shows the average total r.equired charge (ltC) obtained for each dose of diazepam (mg/kg) in individual subjects. The use of charge combines the information conveyed in the individual trade-off curves in Fig. 3 into one number which represents the entire curve. Charge was calculated for each period threshold, summed across currents, and then averaged across replications to yield one value. The data are presented with the posterior VTA site first (TH11) and the LPO site last (TH22).
172 t-Test comparisons were conducted on the total charge values for all drug conditions within each animal. Note that subjects URF62 and F472 were not tested with the higher dose, 5.0 mg/kg, of diazepam. The results of the t-test analysis are shown in Table I. The three seizure-positive animals, TH20, F472, and TH22, all showed significant differences in charge between the vehicle and drug conditions (0 vs 2.5 mg/kg and 0 vs 5.0 mg/kg). In one subject (TH22), the difference between the 2.5 and 5.0 mg/kg doses was also statistically significant. In all of these cases, the total charge required to obtain threshold decreased when diazepam was administered. Among the five seizure-negative placements, only two subjects (TH 11 and TH 19)had significant results. In the case ofTH 11, a significant difference was found between the vehicle and 2.5 mg/kg condition with the mean total charge decreasing in the drug condition, which is the same pattern that was observed in the seizure-positive placements. The second subject, TH 19, showed significant differences between the vehicle and 5.0 mg/kg and between the 2.5 mg/kg and 5.0 mg/kg conditions. However, unlike the other subjects, total charge after 5.0 mg/kg of diazepam was TABLE I t-Test analysis of charge Subject
t-Test comparisons 0 vs 2.5 mg/kg
0 vs 5.0 mg/kg
2.5 vs 5.0 mg/kg
THII
4.518 (5) P < 0.01
4.419 (3) NS
0.112 (4) NS
THI9
1.595 (8) NS
3.500 (9) P < 0.01
3.523 (7) P < 0.01
THI8
3.083 (10) NS
3.073 (10) NS
0.697 (8) NS
URF62
0.403 (5) NS
TH21
2.480 (7) NS
1.632 (8) NS
1.558 (7) NS
TH20
6.724 (8) P < 0.001
6.020 (7) P < 0.001
0.821 (7) NS
F472
11.542 (6) P < 0.001
TH22
14.085 (6) P < 0.001
31.660 (6) P < 0.001
4.895 (6) P < 0.01
Bracketed number refers to degrees of freedom
greater than after either the vehicle or the lower concentration of diazepam.
DISCUSSION
To date, the effects of diazepam on brain-stimulation reward have mainly been studied using traditional rate measures, the results of which have been equivocal. In this study, the interaction between diazepam and the reward system was investigated using a threshold procedure in order to scale out the contribution of performance factors. Diazepam's effect on self-stimulation thresholds appears to be linked to its anti-convulsive action. The group with relatively little or no effect of diazepam on period thresholds were the five subjects exhibiting no signs Of stimulation-induced seizures while the three animals that showed large upward shifts in threshold presented with overt seizure signs that ranged from wet dog shakes, mouth movements, and freezing (TH20) to fully developed motor seizures (F472 and TH22). Two of the five rats in the seizure-negative group showed small effects of diazepam which were not consistent with the changes seen in the seizure-positive group. One subject, THI9, exhibited a significant increase in required charge during the 5.0 mg/kg tests compared with the vehicle and 2.5 mg/kg dose. This corresponds to a decrease in the rewarding value of the stimulation which was not seen in any of the other rats. In addition, the increase in required charge occurred only after the highest dose ofdiazepam while the typical pattern observed was a threshold change at the lowest dose of diazepam with little further contribution at the higher dose. The second rat, TH 11, showed a decrease in total charge in both drug tests when compared to the vehicle condition. This effect corresponds closely to what was seen in the seizure-positive group although the observed changes for TH 11 were somewhat smaller. There were no distinguishing characteristics about this animal other than electrode placement which was the most posterior site of all of the subjects. The seizure-positive group uniformly exhibited large highly significant decreases in charge after 2.5 mg/kg of diazepam was administered. In one case, TH22, a substantial decrease after 5.0 mg/kg of diazepam was also seen. Administration of diazepam resulted in a reduction or elimination of seizures concomitant with the alteration in period thresholds. Diazepam is known to have anti-convulsive properties that suppress motor seizures although electrical afterdischarges are not influenced~9. The concordance between the presence of stimulation-induced seizures
173 and the effectiveness o f d i a z e p a m in attenuating the seizures and altering period thresholds suggests that d i a z e p a m does not directly interact with the reward substrate. Rate m e a s u r e s have been primarily used to examine changes in reward induced by d i a z e p a m . Inconsistent results have been found with rates increasing, decreasing, or unchanging with no clear trend to be seen between studies, animals, currents, electrode sites, and d o s e s tested 1"4'2~ In this study, a cursory examination o f rates also revealed no clear pattern following d i a z e p a m administration. T h e use o f the threshold m e a s u r e dissociated the p e r f o r m a n c e effects o f d i a z e p a m f r o m the a s s e s s m e n t o f brain-stimulation reward after d i a z e p a m challenge. In addition, we investigated the relationship between self-stimulation thresholds and the seizures that often a c c o m p a n y this behavior. D i a z e p a m ' s anti-convulsive actions m a y explain why this c o m p o u n d a p p e a r s to shift self-stimulation thresholds at s o m e sites and is ineffective in others. One possibility is that seizures reduce the rewarding value o f stimulation which is restored with d i a z e p a m ; another is that seizures p o s e a p e r f o r m a n c e obstacle great enough to shift period thresholds to the left. I f so, such effects need to be carefully considered when estimating the rewarding value o f seizure-prone placements.
ACKNOWLEDGEMENTS This w o r k was s u p p o r t e d by N S E R C G r a n t U O 5 1 4 to C.B. and an M R C Research Studentship to T . H . W e would like to t h a n k G e o r g e F o u r i e z o s for his critical r e v i e w o f this paper. REFERENCES 1 Carden, S.E. and Coons, E.E., Diazepam modulated lateral hypothalamic self-stimulation but not stimulation-escape in rats, Brahl Res., 483 (1989) 327-334. 2 Carden, S.E. and Coons, E.E., Diazepam's impact on selfstimulation but not stimulation-escape suggests hedonic modulation, Behav. Neurosci., 104 (1990) 56-61. 3 Caudarella, M., Campbell, K.A. and Milgram, N.W., Differential effects ofdiazepam (valium) on brain stimulation reward sites, Physiol. Biochem. Behav., 16 (1982) 17-21. 4 Caudarella, M., Destrade, C., Cazala, P. and Gauthier, M., Dissociation of limbic structures by pharmacological effects of diazepam on electrical self-stimulation in the mouse, Brabt Res., 302 (1984) 196-200. 5 Edmonds, D.E. and Gallistel, C.R., Parametric analysis of brain stimulation reward in the rat. III. Effect of performance variables
on the reward summation function,J. Comp. PhysioL PsychoL, 87 (1974) 876-883. 6 Edmonds, D.E. and Gallistel, C.R., Reward versus performance in self-stimulation: Electrode-specific effects of ~-methyl-p-tyrosine on reward in the rat, J. Comp. PhysioL PsychoL, 91 (1977) 962-974. 7 Edmonds, D.E., Stellar, J.R. and Gallistel, C.R., Parametric analysis of brain stimulation reward inthe rat. II. Temporal summation in the reward system, J. Comp. Physiol. PsychoL, 87 (1974) 860-869. 8 Gallistel, C.R., Self-stimulation in the rat: quantitative characteristics of the reward pathway, J. Comp. PhysioL PsychoL, 92 (1978) 977-998. 9 Haefely, W.E., Structure and function of the benzodiazepine receptor, Cl,imia, 41 (1987) 389-396. 10 Liebman, J.M., Discriminating between reward and performance: a critical review of intracranial self-stimulation methodology, Neurosci. Biobehav. Rev., 7 (1983) 45-72. 11 Liebman, J.M., Anxiety, anxiolyties and brain stimulation reinforcement, Neurosci. Biobehav. Rev., 9 (1985) 75-86. 12 Miliaressis, E., Rompre, P.-P., Laviolette, P., Philippe, L. and Coulombe, D., The curve-shift paradigm in self-stimulation, Physiol. Behav., 37 (1986) 85-91. 13 Mundl, W.J., A constant-current stimulator, Physiol. Behav., 24 (1980) 991-993. 14 Nomikos, G.G. and Spyraki, C., Effects of ritanserin on the rewarding properties of D-amphetamine, morphine and diazepam revealed by conditioned place preference in rats, Pharmacol. Biochem. Behav., 30 (1988) 853-858. 15 Olds, M.E., Effectiveness of morphine and ineffectiveness of diazepam and phenobarbital on the motivational properties of hypothalamic self stimulation behaviour, Neuropharmacology, 15 (1976) 117-131. 16 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, 2nd Ed., Academic Press, New York, 1986. 17 Porrino, L.J. and Coons, E.E., Effects of GABA receptor blockade on stimulation-induced feeding and self-stimulation, Pharmacol. Biochem. Behavl, 12 (1980) 125-130. 18 Racine, R.J., Modification of seizure activity by electrical stimulation. II. Motor seizure, ElectroencephaL Clin. NeurophysioL, 32 (1972) 281-294. 19 Racine, R., Livingston, K. and Joaquin, A., Effects of procaine hydrochloride, diazepam, and diphenylhydantoin on seizure development in cortical and subcortical structures in rats, ElectroencephaL Clin. Neurophysiol., 38 (1975) 355-365. 20 Reid, L.D., Gibson, W.E., Gledhill, S.M. and Porter, P.B., Anticonvulsant drugs and self-stimulation behavior, J. Comp. PhysioL PsychoL, 57 (1964) 353-356. 21 Robertson, A., Laferriere, A. and Milner, P.M., Treatment with anticonvulsant drugs retards the development of brain-stimulation reward in the prefrontal cortex, PhysioL Behav., 29 (1982) 275-280. 22 Spyraki, C. and Fibiger, H.C., A role for the mesolimbic dopamine system in the reinforcing properties of diazepam. Psychopharmacology, 94 (1988) 133-137. 23 Valenstein, E.S., Problems of measurement and interpretation with reinforcing brain stimulation, Psychol. Rev., 71 (1964) 415 -436. 24 Weinreich, D. and Clark, L.D., Anticonvulsant drugs and selfstimulation rates in rats, Arch. hit. Pharmacodyn., 185 (1970) 269-273.
167
BBR01242
Diazepam alters brain-stimulation reward thresholds in seizure-prone sites T. Harris and C. Bielajew School of Psychology, University of Ottawa, Ottawa, Ontario (Canada)
(Received 8 January 1991) (Revised version received 8 July 1991) (Accepted 21 September 1991) Key words: Benzodiazepine; Diazepam; Brain-stimulation reward; Self-stimulation; Lateral hypothalamus; Ventral tegmental area; Lateral preoptic area; Seizure
Studies of the effect of diazepam and related compounds on the rewarding properties of brain stimulation as measured by response rates have not yielded clear results, with self-stimulation performance reported to be potentiated, diminished, or unchanged folloMng drug administration. In this study, the effect of two doses ofdiazepam (2.5 and 5.0 mg/kg) and its vehicle on self-stimulation thresholds was examined in eight rats with electrode placements scattered along a 4 mm length of the medial forebrain bundle. Stimulation of the lateral preoptie area and the anterior and mid-lateral hypothalamus produced overt seizures. Rate-period curves were generated for a wide range of currents and the resulting period-current trade-off functions were compared across doses. In seizure-prone sites upward shifts in period threshold were observed after 2.5 mg/kg ofdiazepam with little additional increases incurred by the 5.0 mg/kg dose. The majority ofnon-seizure sites showed no effects ofdiazepam upon period threshold. The results suggest that diazepam alters brain-stimulation reward thresholds by suppressing competing seizure activity.
INTRODUCTION The benzodiazepines constitute a class o f m i n o r tranquilizers that have anti-convulsive, sedative, and antianxiety properties 9. Reports o f the effect o f these c o m p o u n d s on brain-stimulation reward in the lateral hypothalamus ( L H ) have not been consistent II. Some studies have found a facilitation o f self-stimulation response rates at low doses (0.5-10 mg/kg) of benzo~ diazepines ~'4'15 while others have found either no effect or a decrease in rates over the same range o f doses 3"4. Likewise with higher doses (10-120 mg/kg), both facilitated and depressed response profiles have been reported despite observed low dose elevations in rates in the same animals ~s. These equivocal results may be attributable to an unpredictable interaction between the sedative actions o f benzodiazepines at high doses and the effects o f brain stimulation, by either promoting rates through suppression o f aversion in mixed placements or reducing rates due to sedation in pure-reward sites.
Correspondence: C. Bielajew, School of Psychology, University of Ottawa, Ottawa, Ontario, Canada, KIN 6N5.
The influence o f benzodiazepines on extrahypothalamic placements has not been extensively studied. Caudarella et al. 4 found that low doses o f diazepam facilitated L H response rates while dorsal hippocampal rates were attenuated. In the prefrontal cortex, response rates were unaffected if d i a z e p a m was administered to animals following acquisition o f self-stimulation behaviour at. A major d r a w b a c k o f these studies is the pervasive use o f response rate as an index o f reward; the assumption that drug-induced changes in rate indicate a shift in reward is unfounded. T h e performance enhancing or debilitating effects o f some drugs may mask the 'true' reward effects 5,6,1~ T h e threshold method for assessing changes in self-stimulation controls and minimizes the contribution o f performance factors. Typically, current is held constant while the frequency o f pulses, or its reciprocal, period, is systematically varied to generate what has been termed a 'reward summation' function 5, that is, a r a t e - f r e q u e n c y or r a t e - p e r i o d curve. Threshold is calculated by interpolating the frequency or period corresponding to a criterion response rate. Because absolute response rates are ignored, the problems ofinterprctation due to non-motivational fac-
168 tors are eliminated and the reliability of the data strengthened. Very few studies have evaluated threshold changes in self-stimulation following benzodiazepine administration. In the prefrontal cortex, current thresholds are unaltered by diazepam challenge2% while in the LH, the frequency thresholds appear to decrease in subjects with pure reward placements; this finding is less consistent in subjects with mixed reward-aversion sitesL To our knowledge, increased current or frequency thresholds have not been observed, suggesting that the benzodiazepines, when effective, amplify the rewarding properties of electrical stimulation. This notion is strengthened by the finding that administration ofpicrotoxin, a benzodiazepine antagonist, results in increased frequency thresholds, implying a facilitation of reward iT. Finally, diazepam has been reliably shown to produce a significant conditioned place preference ~4"22, consistent with the notion that diazepam has rewarding effects. However, the question still remains as to why the benzodiazepines are not effective in altering self-stimulation thresholds at all sites, and in all subjects at the same site. One suggestion is that diazepam's rewardenhancing effects are reduced in placements with aversive properties ~'2. If so, this phenomenon seems to be restricted to the hypothalamic region as reward thresholds in sites not usually associated with aversion (e.g. prefrontal cortex) are basically unaffected by diazepam 2~. A second possibility, one that emerged from our pilot data, is that the anti-convulsive action of benzodiazepines suppresses the development or appearance of motor seizures that inhibit the behavioral expression ofthe rewarding effects of the stimulation; thus, in those animals not experiencing seizures, self-stimulation thresholds would be unaffected by the anti-convulsant property of the drugs. This interpretation is difficult to assess in earlier studies because seizure activity was not reported. Thus, the present study was initiated to determine if threshold changes were correlated with the presence of seizures. Electrodes were implanted along a wide posteroanterior axis of the medial forebrain bundle from the ventral tegmental area (VTA) to the lateral preoptic area (LPO) to include sites where the probability of observing stimulation-induced seizures was high.
MATERIALS AND METHODS
Subjects attd surgery Eightmale Long-Evans rats (Charles River Laboratories) were individually housed in plastic cages under
a 12 I1 light/dark cycle with ad libitum access to food and water. At the time of surgery, the individual weights ranged from 300 to 440 g. The rats were anesthetized with sodium pentobarbital (65 mg/kg i.p.) and xylazine (5 mg/kg i.m.). A subcutaneous injection of atropine sulfate was administered to reduce respiratory distress. Standard stereotaxic procedures were used to implant the electrodes in one of three sites using a fiat-skull orientation ~6. The three VTA subjects had coordinates ranging from 4.3 to 4.8 mm posterior to bregma, 0.9 to 1.2 mm lateral to the mid-sagittal suture, and 8.4 to 8.8 mm below the skull surface reading at bregma; the four LH subjects had coordinates ranging from 2.0 to 3.0 mm posterior, 1.6 to 1.7 mm lateral, and 8.2 to 8.4 mm ventral; the one LPO subject had coordinates of 0.6 mm posterior, 1.3 mm lateral, and 8.2 mm ventral. The electrodes were fashioned from stainless-steel wire, 250 pm in diameter, and insulated with Epoxylite to the flat tips. The indifferent electrode consisted of a 2-56 stainless-steel machine screw with a stainless-steel flexible wire soldered to it; the wire was wrapped around the four stainless-steel skull screws. The entire assembly was anchored to the skull screws by dental cement.
Apparatus The testing chamber was a 28 cm long x 38 cm wide • 44 cm tall wooden box with a Plexiglas front. A rodent lever protruded from a side wall and was positioned 3 em above the floor. Stimulation was delivered by a constant-current amplifier ~3 and an integrated circuit pulse generator, built in-house. The current was continuously monitored on an oscilloscope by reading the voltage drop across a 1 kf~ precision resistor in series with the rat. Each lever depression delivered a 500 ms train of rectangular monophasic cathodal pulses of 100#s duration. The pulse period and current were varied as described below.
Drugs Diazepam (Hoffman-LaRoche) was obtained in injection ampules at a concentration of 5 mg/ml. The vehicle solution was prepared from 414 mg propylene glycol and 80 mg ethanol per ml of saline, having a pH of 7.3, and was filtered before each injection.
Screen#lg and stabilization Animals were allowed to recover from surgery for at least 3 days before screening for self-stimulation began. Standard operant shaping procedures were used to train the rats to bar-press for electrical stimulation.
169 Once reliable performance was observed, the onset of each 60 s trial was signalled by 5 stimulation primes with current and period identical to the 60 s trial stimulation; the inter-prime interval was 1.0 s. The current and period were adjusted to produce rates of responding greater than 50 per minute. During stability sessions, an ascending series of periods was introduced with the initial period sufficient to produce high rates of responding and each successive period a 0.10 common log unit step above the previous one. The series was continued until the rate of responding in a 60 s trial was 5 bar presses or fewer. This procedure was applied to 6 or 7 currents which were spaced roughly equally from 80 ILA to 1000/~A. Rate was plotted against period for each current to determine the threshold, which was defined as that period corresponding to 25 responses and was interpolated from the rate-period functions shown in Fig. 1. From this family of curves, a period-current trade-off function was generated. Two measures of stability were used to determine when a subject was ready to begin drug tests. Within a session, a current, usually 200 ltA, and its associated period threshold was designated as the baseline measure. If the baseline threshold measured at the beginning and at the end of the session differed by no more than 0.10 common log units, the data from that session were retained; if greater, the session was discarded. The second stability measure was a between-
foe 9r
I00 8C ~
.~
160
250
5OO \ LA I000
7C
session criterion. The standard error of the mean associated with the average period threshold across sessions could not exceed 10~ of the mean value. Replications of the full period-current trade-off function were continued until the majority of the average period thresholds met this criterion; typically this required from 4-6 sessions, with a single period-current trade-off function generated per session. A session lasted approximately 2 h. Once the between-session stability was established, drug testing began.
Behavioural testhlg Six of the eight subjects were tested with two doses of diazepam, 2.5 mg/kg and 5.0mg/kg, while the remaining two subjects (URF62 and F472) were tested with only the 2.5 mg/kg dose. All subjects received vehicle injections. During test sessions, the incidence of stimulation-induced seizures or signs indicating the development of seizures was recorded. Drug and vehicle days were interdigitated so that drug tests were conducted no less than 48 h apart. On those days, the two doses ofdiazepam were assigned randomly. Within each session, the currents were also tested on a random schedule to yield a full period-current trade-off curve. The between-session criterion mentioned above was imposed during this phase of the study; thus, 4-6 replications of the entire period trade-off function were collected under each drug condition. The subjects were injected intraperitoneally approximately 10 min before testing began. This delay allowed the overt signs of the sedation induced by diazepam to subside. The duration of the test sessions was roughly one hour, during which time, stable baseline measures were usually observed.
Histology Subjects were perfused transcardially with a solution of 0.9~o saline followed by 10~ formalin. The brains were removed from the skull and stored in 10~o formalin for at least 48 h before sectioning and staining with Cresyl violet. The Paxinos and Watson atlas ~6was used to verify the location of the electrode tips.
i
\ 5O
... 2C
RESULTS 4.0
""
~3
~ao
I~e
zs.~
3s.
63.1
I00.0
PERIOD THRESHOLD (msec)
Fig. I. An example of period threshold determination from r a t e - period functions collected at different currents. A constant criterion of 25 responses/min was used to evaluate period threshold as indicated by the horizontal dashed lines. The corresponding thresholds which are indicated by the downward arrows were used to derive the period-current trade-off functions presented in Fig. 3.
Histology Tracings of the atlas plates ~6 corresponding to the sections containing the electrode tips are shown in Fig. 2. Subject TH22's electrode was located within the limits of the LPO, slightly ventral and medial to the substantia innominata. Of the four LH electrodes, three of the tips were found at different depths in the same
170
TH22
F472 oURF62 mTH21 ATH20
THIS
THI9
THII
Fig. 2. Tracings from atlas plates (Paxinos and Watson, 1986) that best correspond to the sections containing the electrode tips. The anteroposterior level is shown at the bottom of each tracing. The alphanumeric listing on the right of each section identifies the subject. Electrode placements ranged from the posterior LPO to the posterior VTA.
anteroposterior level. One tip (TH20) was positioned between the dorsomedial hypothalamic nucleus and the perifornical area, just dorsal to the level of the fornix. Another (TH21) had a similar placement but was marginally more dorsal and lateral, just below the subincertal nucleus. The last tip located at this level (URF62) was in the zona incerta just lateral to the mammillothalamic tract. Subject F472's electrode was situated far more anterior to the other LH placements with the tip in the most lateral part ofthe LH just dorsal to the magnocellular preoptic area. The three VTA electrodes were found at different anteroposterior levels. The most anterior (TH18) was at the dorsal border of the VTA, slightly lateral to the fasciculus retroflexus. The second (TH 19) was just ventral and medial to the medial lemniscus, and the most posterior VTA placement, (TH 11), was positioned at the ventral border of the VTA encroaching on the medial lemniscus and paranigral nucleus. Dntg Tests
The effect of diazepam upon brain stimulation reward is shown in Fig. 3.
The subjects are presented according to their anteroposterior electrode placement, from left to right beginning at the top, so that the data from the subject with the most posterior electrode placement (TH 11) appears in the top left box and the data from the subject with the most anterior electrode placement (TH22) appears in the bottom right box. The individual curves represent the average period thresholds plotted against current (period-current trade-off function) that were obtained under each drug condition with vehicle tests distinguished by filled circles and solid lines, 2.5 mg/kg tests distinguished by filled circles and dashed lines, and 5.0 mg/kg tests distinguished by open circles and solid lines. Seizures which ranged from head and mouth movements with wet dog shakes to fully expressed motor seizures with rearing and falling were observed repeatedly during vehicle testing in TH22, F472, and TH20. Administration of diazepam eliminated or reduced the severity of the seizures. The remaining five animals showed no overt seizure signs. Stimulation of the five seizure-negative sites resulted in little or no effect of diazepam upon self-stimulation thresholds at either the 2.5 mg/kg or 5.0 mg/kg dose. However, the three seizure-positive placements showed a striking increase in period thresholds, ranging from 0.1 to 0.4 common log units, following diazepam challenge. In these three subjects, threshold shifts were observed with the 2.5 mg/kg dose; the higher dose yielded little or no additional increase in period thresholds. The total charge was calculated for each of the individual period-current trade-off functions using the equation Q = INd
where Q is the charge in ILC, I is the current in/tA, N is the number of pulses in the stimulation train, and d is the pulse duration in seconds 8. The charge value across currents should remain constant if there is an equal reciprocity between period and current; if so, the slope of the period-current trade-off function that relates these two variables, when plotted using logarithmic (base 10) scales, should be 1. Although there is a wide range of slopes across subjects in this study, the slopes of the trade-off functions within subjects are similar, implying that equivalent changes are occurring at each current across drug tests in individual subjects. The use of charge simplifies the data analysis by combining the information conveyed in the trade-off curves into one number which represents the entire curve. Therefore, for every one of the 4-6 replications of each
171 158.5 1259
HII
I000
f
79.4
THI9 -5.3
THIS
URF62 -31
TH20 -5.1
F472
TH22 -0.6
6~1.1 50.1 398 31.6 25.1 20.0 "~
158 126 I00
7.9 6l]
::I:: 4o ~ 3.2 1~5.9 RO0.O
79.4
TH21 -3.1
;e
-I.8
.~./.
seizures
63.1
seizures
seizures s"
I~.
39.8
_ 4 . "~
p.-
31.6 25.i
S/
20.0 158
se~
12.6 I00
T.9 6.3 5.0 4.0 3240
~;
l ,GO ' '~0 ' z~O ' 4;0
~iO ' '0;0
9 6~ ',oo " ,6o =~o " ,Go" 6~o ,ooo
63
CURRENT
,oo ' ,~o " =~o
,Go
6~o " ,o'oo
9 s'3 ' ,6o"i~o
=5o',bo'e~o'lo'oo
(p-A)
Fig. 3. The average period-current trade-off functions determined for all drug conditions with period threshold plotted against current. Each line is based on 4-6 replications. In the upper left corner of each box is the subject's identification number and below that, the anteroposterior coordinate for that subject's electrode placement. Filled circles joined by a solid line represent the vehicle condition; filled circles joined by a dashed line represent the 2.5 mg/kg diazepam condition; open circles joined by a solid line represent the 5.0 mg/kg diazepam condition.
trade-off function shown in Fig. 3, Q was calculated for every data point, that is, each period threshold value that contributed to the curve. The charge values were
~7
summed across the currents.to yield a total charge for each replication which was then averaged to yield one representative value for that test condition (Fig. 4).
~,b,c
a-O vs 2.Srng/kg p
46 btc
.7-
.Z.J.
J. _z. dr..
E 3 t.t.
0 2.5 5.0
THII
2~ 5,0
THI9
0 2.5 S.O
0 2-5
0 2.5 SD
THI8
URF62
TH21
Z.5 5.0
TH20
0 25
F472
1
0 2~5S0
TH22
DOSE OF DIAZEPAM (mg/kg)
Fig. 4. The graph shows the average total r.equired charge (ltC) obtained for each dose of diazepam (mg/kg) in individual subjects. The use of charge combines the information conveyed in the individual trade-off curves in Fig. 3 into one number which represents the entire curve. Charge was calculated for each period threshold, summed across currents, and then averaged across replications to yield one value. The data are presented with the posterior VTA site first (TH11) and the LPO site last (TH22).
172 t-Test comparisons were conducted on the total charge values for all drug conditions within each animal. Note that subjects URF62 and F472 were not tested with the higher dose, 5.0 mg/kg, of diazepam. The results of the t-test analysis are shown in Table I. The three seizure-positive animals, TH20, F472, and TH22, all showed significant differences in charge between the vehicle and drug conditions (0 vs 2.5 mg/kg and 0 vs 5.0 mg/kg). In one subject (TH22), the difference between the 2.5 and 5.0 mg/kg doses was also statistically significant. In all of these cases, the total charge required to obtain threshold decreased when diazepam was administered. Among the five seizure-negative placements, only two subjects (TH 11 and TH 19)had significant results. In the case ofTH 11, a significant difference was found between the vehicle and 2.5 mg/kg condition with the mean total charge decreasing in the drug condition, which is the same pattern that was observed in the seizure-positive placements. The second subject, TH 19, showed significant differences between the vehicle and 5.0 mg/kg and between the 2.5 mg/kg and 5.0 mg/kg conditions. However, unlike the other subjects, total charge after 5.0 mg/kg of diazepam was TABLE I t-Test analysis of charge Subject
t-Test comparisons 0 vs 2.5 mg/kg
0 vs 5.0 mg/kg
2.5 vs 5.0 mg/kg
THII
4.518 (5) P < 0.01
4.419 (3) NS
0.112 (4) NS
THI9
1.595 (8) NS
3.500 (9) P < 0.01
3.523 (7) P < 0.01
THI8
3.083 (10) NS
3.073 (10) NS
0.697 (8) NS
URF62
0.403 (5) NS
TH21
2.480 (7) NS
1.632 (8) NS
1.558 (7) NS
TH20
6.724 (8) P < 0.001
6.020 (7) P < 0.001
0.821 (7) NS
F472
11.542 (6) P < 0.001
TH22
14.085 (6) P < 0.001
31.660 (6) P < 0.001
4.895 (6) P < 0.01
Bracketed number refers to degrees of freedom
greater than after either the vehicle or the lower concentration of diazepam.
DISCUSSION
To date, the effects of diazepam on brain-stimulation reward have mainly been studied using traditional rate measures, the results of which have been equivocal. In this study, the interaction between diazepam and the reward system was investigated using a threshold procedure in order to scale out the contribution of performance factors. Diazepam's effect on self-stimulation thresholds appears to be linked to its anti-convulsive action. The group with relatively little or no effect of diazepam on period thresholds were the five subjects exhibiting no signs Of stimulation-induced seizures while the three animals that showed large upward shifts in threshold presented with overt seizure signs that ranged from wet dog shakes, mouth movements, and freezing (TH20) to fully developed motor seizures (F472 and TH22). Two of the five rats in the seizure-negative group showed small effects of diazepam which were not consistent with the changes seen in the seizure-positive group. One subject, THI9, exhibited a significant increase in required charge during the 5.0 mg/kg tests compared with the vehicle and 2.5 mg/kg dose. This corresponds to a decrease in the rewarding value of the stimulation which was not seen in any of the other rats. In addition, the increase in required charge occurred only after the highest dose ofdiazepam while the typical pattern observed was a threshold change at the lowest dose of diazepam with little further contribution at the higher dose. The second rat, TH 11, showed a decrease in total charge in both drug tests when compared to the vehicle condition. This effect corresponds closely to what was seen in the seizure-positive group although the observed changes for TH 11 were somewhat smaller. There were no distinguishing characteristics about this animal other than electrode placement which was the most posterior site of all of the subjects. The seizure-positive group uniformly exhibited large highly significant decreases in charge after 2.5 mg/kg of diazepam was administered. In one case, TH22, a substantial decrease after 5.0 mg/kg of diazepam was also seen. Administration of diazepam resulted in a reduction or elimination of seizures concomitant with the alteration in period thresholds. Diazepam is known to have anti-convulsive properties that suppress motor seizures although electrical afterdischarges are not influenced~9. The concordance between the presence of stimulation-induced seizures
173 and the effectiveness o f d i a z e p a m in attenuating the seizures and altering period thresholds suggests that d i a z e p a m does not directly interact with the reward substrate. Rate m e a s u r e s have been primarily used to examine changes in reward induced by d i a z e p a m . Inconsistent results have been found with rates increasing, decreasing, or unchanging with no clear trend to be seen between studies, animals, currents, electrode sites, and d o s e s tested 1"4'2~ In this study, a cursory examination o f rates also revealed no clear pattern following d i a z e p a m administration. T h e use o f the threshold m e a s u r e dissociated the p e r f o r m a n c e effects o f d i a z e p a m f r o m the a s s e s s m e n t o f brain-stimulation reward after d i a z e p a m challenge. In addition, we investigated the relationship between self-stimulation thresholds and the seizures that often a c c o m p a n y this behavior. D i a z e p a m ' s anti-convulsive actions m a y explain why this c o m p o u n d a p p e a r s to shift self-stimulation thresholds at s o m e sites and is ineffective in others. One possibility is that seizures reduce the rewarding value o f stimulation which is restored with d i a z e p a m ; another is that seizures p o s e a p e r f o r m a n c e obstacle great enough to shift period thresholds to the left. I f so, such effects need to be carefully considered when estimating the rewarding value o f seizure-prone placements.
ACKNOWLEDGEMENTS This w o r k was s u p p o r t e d by N S E R C G r a n t U O 5 1 4 to C.B. and an M R C Research Studentship to T . H . W e would like to t h a n k G e o r g e F o u r i e z o s for his critical r e v i e w o f this paper. REFERENCES 1 Carden, S.E. and Coons, E.E., Diazepam modulated lateral hypothalamic self-stimulation but not stimulation-escape in rats, Brahl Res., 483 (1989) 327-334. 2 Carden, S.E. and Coons, E.E., Diazepam's impact on selfstimulation but not stimulation-escape suggests hedonic modulation, Behav. Neurosci., 104 (1990) 56-61. 3 Caudarella, M., Campbell, K.A. and Milgram, N.W., Differential effects ofdiazepam (valium) on brain stimulation reward sites, Physiol. Biochem. Behav., 16 (1982) 17-21. 4 Caudarella, M., Destrade, C., Cazala, P. and Gauthier, M., Dissociation of limbic structures by pharmacological effects of diazepam on electrical self-stimulation in the mouse, Brabt Res., 302 (1984) 196-200. 5 Edmonds, D.E. and Gallistel, C.R., Parametric analysis of brain stimulation reward in the rat. III. Effect of performance variables
on the reward summation function,J. Comp. PhysioL PsychoL, 87 (1974) 876-883. 6 Edmonds, D.E. and Gallistel, C.R., Reward versus performance in self-stimulation: Electrode-specific effects of ~-methyl-p-tyrosine on reward in the rat, J. Comp. PhysioL PsychoL, 91 (1977) 962-974. 7 Edmonds, D.E., Stellar, J.R. and Gallistel, C.R., Parametric analysis of brain stimulation reward inthe rat. II. Temporal summation in the reward system, J. Comp. Physiol. PsychoL, 87 (1974) 860-869. 8 Gallistel, C.R., Self-stimulation in the rat: quantitative characteristics of the reward pathway, J. Comp. PhysioL PsychoL, 92 (1978) 977-998. 9 Haefely, W.E., Structure and function of the benzodiazepine receptor, Cl,imia, 41 (1987) 389-396. 10 Liebman, J.M., Discriminating between reward and performance: a critical review of intracranial self-stimulation methodology, Neurosci. Biobehav. Rev., 7 (1983) 45-72. 11 Liebman, J.M., Anxiety, anxiolyties and brain stimulation reinforcement, Neurosci. Biobehav. Rev., 9 (1985) 75-86. 12 Miliaressis, E., Rompre, P.-P., Laviolette, P., Philippe, L. and Coulombe, D., The curve-shift paradigm in self-stimulation, Physiol. Behav., 37 (1986) 85-91. 13 Mundl, W.J., A constant-current stimulator, Physiol. Behav., 24 (1980) 991-993. 14 Nomikos, G.G. and Spyraki, C., Effects of ritanserin on the rewarding properties of D-amphetamine, morphine and diazepam revealed by conditioned place preference in rats, Pharmacol. Biochem. Behav., 30 (1988) 853-858. 15 Olds, M.E., Effectiveness of morphine and ineffectiveness of diazepam and phenobarbital on the motivational properties of hypothalamic self stimulation behaviour, Neuropharmacology, 15 (1976) 117-131. 16 Paxinos, G. and Watson, C., The Rat Brain in Stereotaxic Coordinates, 2nd Ed., Academic Press, New York, 1986. 17 Porrino, L.J. and Coons, E.E., Effects of GABA receptor blockade on stimulation-induced feeding and self-stimulation, Pharmacol. Biochem. Behavl, 12 (1980) 125-130. 18 Racine, R.J., Modification of seizure activity by electrical stimulation. II. Motor seizure, ElectroencephaL Clin. NeurophysioL, 32 (1972) 281-294. 19 Racine, R., Livingston, K. and Joaquin, A., Effects of procaine hydrochloride, diazepam, and diphenylhydantoin on seizure development in cortical and subcortical structures in rats, ElectroencephaL Clin. Neurophysiol., 38 (1975) 355-365. 20 Reid, L.D., Gibson, W.E., Gledhill, S.M. and Porter, P.B., Anticonvulsant drugs and self-stimulation behavior, J. Comp. PhysioL PsychoL, 57 (1964) 353-356. 21 Robertson, A., Laferriere, A. and Milner, P.M., Treatment with anticonvulsant drugs retards the development of brain-stimulation reward in the prefrontal cortex, PhysioL Behav., 29 (1982) 275-280. 22 Spyraki, C. and Fibiger, H.C., A role for the mesolimbic dopamine system in the reinforcing properties of diazepam. Psychopharmacology, 94 (1988) 133-137. 23 Valenstein, E.S., Problems of measurement and interpretation with reinforcing brain stimulation, Psychol. Rev., 71 (1964) 415 -436. 24 Weinreich, D. and Clark, L.D., Anticonvulsant drugs and selfstimulation rates in rats, Arch. hit. Pharmacodyn., 185 (1970) 269-273.